Computer method and apparatus for online process identification

Abstract

A computer method and apparatus of online automated model identification of multivariable processes is disclosed. The method and apparatus carries out automatically all the four basic steps of industrial process identification: 1) identification test signal design and generation, 2) identification plant test, 3) model identification and 4) model validation. During the automated plant test, process models will be automatically generated at a given time interval, for example, every hour, or on demand; the ongoing test can be automatically adjusted to meet the process constraints and to improve the data quality. Plant test can be in open loop operation, closed-loop operations or partly open loop and partly closed-loop. In a (partial) closed-loop plant test, any type of controller can be used which include proportional-integral-derivative (PID) controllers and any industrial model predictive controller (MPC). The obtained process models can be used as the model in advanced process controllers such as model predictive control (MPC) and linear robust control; they can also be used as inferential models or soft sensors in prediction product qualities. The apparatus can be used in new MPC controller commissioning as well as in MPC controller maintenance.

1 . Computer apparatus that performs online and automatic identification of dynamic process models for use in model predictive control (MPC) and other advanced process control (APC) compromising: 
 a testing device that carries out the plant test automatically and collects process data; and    a model identification device that carries out model identification and validation and adjust the test parameters automatically using collected process data.    
   
   
       2 . Computer apparatus as claimed in  claim 1  wherein the testing device and the model identification device are interconnected seamlessly so that the whole identification procedure including plant test and model identification (computation) is done online and automatically.  
   
   
       3 . Computer apparatus as claimed in  claim 1  wherein the testing device uses automated multivariable plant test that move many or all MVs simultaneously.  
   
   
       4 . Computer apparatus as claimed in  claim 1  wherein the testing device uses closed-loop control during the test in order to reduce disturbance to unit operation and the controller can be of any type that includes regulatory PID controllers, any MPC controller, or the combination of the two types.  
   
   
       5 . Computer apparatus as claimed in  claim 1  wherein the testing device will continue testing when the MPC controller is taking control actions.  
   
   
       6 . Computer apparatus as claimed in  claim 1  wherein the testing device uses test signals that are related to unit time to steady state. The planned test time is related to the process time to steady state and the number of MVs.  
   
   
       7 . Computer apparatus as claimed in  claim 1  wherein the testing device uses GBN (generalised binary noise) signals in the plant test for the identification of linear models. A small white noise signal is added to each GBN signal to improve its information content. Normally, the test signals are not correlated by design. However, for certain ill-conditioned processes such as high purity distillation columns, strongly correlated test signals will be used for some MVs.  
   
   
       8 . Computer apparatus as claimed in  claim 1  wherein the testing device can control open loop CVs by move the mean values of some open loop MVs in order to reduce process disturbance. The testing devise can reduce MV step sizes in order to reduce process disturbance.  
   
   
       9 . Computer apparatus as claimed in  claim 1  wherein the testing device adjusts test signal step size and test signal switch time during the plant test in order to improve model quality.  
   
   
       10 . Computer apparatus as claimed in  claim 1  wherein the testing device uses specially designed graphic user interfaces (GUIs) for test monitoring. The MV Window shows both past MV movements and future planed movements. The CV Window shows past CV movements and it will also show predicted CV movements if a model is identified and available.  
   
   
       11 . Computer apparatus as claimed in  claim 1  wherein the model identification device uses the ASYM method in model parameter estimation and order selection.  
   
   
       12 . Computer apparatus as claimed in  claim 1  wherein the model identification device calculates current upper bounds and future upper bounds of model errors.  
   
   
       13 . Computer apparatus as claimed in  claim 1  wherein the calculated upper error bounds are used to grade model qualities, to adjusting the ongoing plant test and to decide the stop time of the plant test.  
   
   
       14 . Computer apparatus as claimed in  claim 1  wherein the model identification device uses truly closed-loop data for identification, that is, the data contain feedback actions of the PID or MPC controller and the test signal.  
   
   
       15 . Computer apparatus as claimed in  claim 1  wherein the model identification device uses the Expectation Matrix in model identification in order to improve computation speed and model quality.  
   
   
       16 . Computer apparatus as claimed in  claim 1  wherein the model identification device estimates process delays (dead times) in order to improve model quality.